MASTER NOTES: Ryan QS and Expected Wins

A year or so ago in this space, I fulminated at length about how the wins categories bites heinie and how fantasy baseball would be much better off if we went to the “Ryan Quality Start,” which is seven innings with three earned runs or fewer allowed.

The fantasy baseball world was inspired to a mass movement of ignoring me and leaving things exactly as they were. But absolute and total humiliating rejection has never stopped me, so I’ve always kept the idea of the “RQS,” as I’ve come to call it, in the back of my mind, along with old Astronomy 100 formulas, where I left the keys to my 1969 Datsun 510, and so on.

The point is that I was thinking about RQS the other day and it suddenly occurred to me that one thing I liked about it as a replacement for wins as a category is that there are roughly as many RQS as pitcher wins. That got me to wondering if RQS could serve as an indicator of wins.

I was particularly curious if there are many pitchers who have been getting lots of wins without rolling up a lot of RQS, in which case we could infer an overabundance of good fortune, wins-wise, and a potential balancing shortfall of wins to come later in the season.

On the other hand, we might find pitchers who have been getting lots of RQS without getting their share of wins. And we might reasonably expect those pitchers' luck to come around and therefore for them to collect some extra wins later in the season.

The first thing I did was to collect all the 2016 pitcher-start data from BaseballHQ.com’s PQS charts. You can download the entire season for each league as an Excel-readable file, and I did. There were 2,460 pitcher-starts in 2016 through July 3.

I first looked at some big-picture questions. Not “what is the meaning of life” type questions, just basic numbers about Quality Starts (QS), Ryan Quality Starts (RQS, as noted) and wins. So here’s the basic skinny:

  • 48% of all starts are QS.
  • 22% of all starts are RQS (all RQS are also QS).
  • 55% of all QS resulted in wins.
  • 64% of all RQS resulted in wins (the same win percentage as PQS 4-5).
  • 25% of non-RQS starts resulted in wins.

Next, I looked at the 181 individual pitchers with at least five starts this year. Only five had no QS, but 34 had no RQS—the first hint that RQS was separating the wheat from the chaff (and, unlike the old journalists’ joke about newsroom editors, keeping the wheat).

Next, a ranking of percentages further connected higher RQS to better starters. The list of guys with at least 70% QS% (percentage of starts that were QS) included 27 names. Many of the elite guys made the list at that level, but the list also had James Shields, Matt Wisler and Colby Lewis.

By contrast, the list of pitchers at 70% RQS ends after Clayton Kershaw and Chris Sale. Extending down to 50% RQS still only gets 14 names, adding Johnny Cueto, Tanner Roark, Jon Lester, Max Scherzer, Corey Kluber, Jake Arrieta, Julio Teheran, Aaron Sanchez, Noah Syndergaard, Steven Wright, Trevor Bauer and Felix Hernandez.

It seems again that with RQS, the cream rises to the top, to employ another agricultural cliché. These guys are $20+ pitchers in most leagues, and even without any other analysis, I think I'll be targeting Bauer and Teheran.

But there was one other analytical step I wanted to take: Given that 64% of RQS are wins, I thought any SP should have had wins in about that percentage of his RQS. And pitchers should have had wins in only about 25% of their non-RQS starts.

This raises two questions:

1. Sorry, what? I dozed off; and

2. Who were the outliers?

Among the aforementioned top RQS pitchers, those underperforming the 64% expected win percentage included Roark (50%), Kluber (56%) and Teheran (33%). Madison Bumgarner, who actually does not have a stellar RQS%, nonetheless has been likewise unfortunate in winning only three of his seven RQS.

The top wins-getters in non-RQS starts included Rich Hill (60%), Josh Tomlin (60%), Danny Salazar (67%), Jordan Zimmermann (70%) and Stephen Strasburg (73%).

Remember, the overall percentage of wins in non-RQS starts is just 25%, so it seems these guys have been enjoying some good fortune that might not be dependable. Indeed, we’ve seen Zimmermann start to slide overall since an excellent 8-for-10 win streak to open the season (despite only four RQS—six of his eight wins were non-RQS).

I ended up with two takeaways from this research effort:

First, while this concept needs more detailed examination, it looks like it might have potential for identifying pitchers getting too few wins or too many on a game-by-game basis.

And second, it looks like it has the potential to help identify underrated good starting pitchers in general, like Roark, Teheran, and Sanchez, and to identify possibly overrated guys like Hill, Tomlin and Zimmermann.

Of course there are obvious external factors that clearly affect win totals. Teheran’s ATL teammates produce barely 3.5 Runs Per Game (R/G), and NYM and PHI are also below 4.0 R/G. Meanwhile, BOS produces 5.7 (making David Price’s league average 64% RQS W% look low) and CHC, STL, COL and BAL are all over 5.0 R/G.

And bullpens vary widely in caliber as well. The MLB average for inherited runners scoring is 32%, but that’s in a range from 15% (HOU) to 44% (NYY, and ain’t that a surprise). CIN has lost 21 games in relief (TOR and ATL close behind at 19) while KC has lost just five, with CHC, BAL, STL, SD and DET all under 10.

Finally, if you aren’t convinced or intrigued by back-of-the envelope research like this, remember that BaseballHQ.com now has a relatively new “Expected Wins” metric in its toolkit for starting pitchers. Developed by Matt Cederholm, xW adapts the Bill James “Pythagorean Theorem,” which determines a team’s expected wins by using its runs scored and allowed, and applies it to SPs. It’s a peach.


Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.